A high percentage of IT and business executives are expected to invest extensively on AI in the next few years. And with more organizations adopting AI systems, system quality-related risks increase. Hence, testing and tooling strategies for AI systems have to be in place to ensure quality. This session will then look into a typical software testing lifecycle and see that it can still be used as an approach for simple AI system implementations. We’ll look into our recent testing approach for an IVR chatbot where we can still leverage the standard testing process. We’ll see that understanding the intents and utterances of the chatbot system is important to ensure that we define various scenarios and conditions on how users will use the chatbot. And given the variety of scenarios and conditions to factor in, we’ll see an example of a test automation implemented for a shorter test cycle and good coverage. At the same time, we’ll also see how a different approach can be implemented for more complex AI system implementations such as those with machine learning.